SShogo Bench

MiniMax-M2.5

Prior MiniMax coding model for agent workflows, office edits, and automation

Shogo Agent Index

Estimated#74 of 92
28.4range 19.5–37.2 · 19% of weight measured
Agentic coding
32
23% measured
Long-horizon autonomy
34
30% measured
Knowledge-work agents
24
estimated
Computer use
17
estimated
Security & context learning
27
50% measured

Price

Input
Free/M
Output
Free/M
Cache read
Free/M
3:1 blend
Free/M
Agent-effective
Free/M

Cheapest host for agents: Venice at $0.11/M.

Agent-effective = 20 input tokens per output token, 85% cache hits. Why.

Agentic benchmarks

BenchmarkScoreNormalisedIndex weightHarness / source
SWE-bench Verifiedest.8%predicted 63
Terminal-Bench 2.042.7%8%cchuter
DeepSWEest.6%predicted 21
FrontierSWEest.5%predicted 3
CursorBenchest.4%predicted 25
GSO-Benchest.4%predicted 7
METR Time Horizonsest.10%predicted 62
Vending-Bench 2$-236%source
PostTrainBenchest.4%predicted 16
APEX-Agentsest.8%predicted 22
GDPvalest.7%predicted 28
Remote Labor Indexest.5%predicted 1
DeepResearch Benchest.5%predicted 44
OSWorld 2.0est.6%predicted 2
OSWorldest.4%predicted 39
Cybenchest.5%predicted 43
CL-bench11.4%5%source

Other benchmarks

BenchmarkCategoryScoreSource
ARC-AGIReasoning63.7%source
ARC-AGI-2Reasoning4.9%source
ProofBenchMath4.0%source
WebDev ArenaCoding1384source

API hosts

Live prices and uptime for every host serving MiniMax-M2.5 via OpenRouter.

HostInputOutputCache readAgent $/MUptime 24hQuant
Venice$0.27$0.95$0.030$0.1199.8%—
GMICloud$0.30$1.20$0.030$0.1298.5%fp8
AtlasCloud$0.29$1.20$0.060$0.15100.0%fp8
DigitalOcean$0.30$1.20$0.060$0.1598.5%—